深度学习：原理与应用实践pdf/txt下载_在线阅读全文

编辑推荐

本书全面、系统地介绍深度学习相关的技术，所有章节均附有源程序，所有实验读者均可重现，具有高度的可操作性和实用性。

内容简介

本书全面、系统地介绍深度学习相关的技术，包括人工神经网络，卷积神经网络，深度学习平台及源代码分析，深度学习入门与进阶，深度学习高级实践，所有章节均附有源程序，所有实验读者均可重现，具有高度的可操作性和实用性。通过学习本书，研究人员、深度学习爱好者，能够在3 个月内，系统掌握深度学习相关的理论和技术。

作者简介

　　　　张重生，博士，教授，硕士生导师，河南大学大数据研究中心、大数据团队带头人。研究领域为大数据分析、深度学习、数据挖掘、数据库、数据流（实时数据分析）。
　　　　博士毕业于 INRIA,France(法国国家信息与自动化研究所)，获得优秀博士论文荣誉。2010年08月至2011年3月，在美国加州大学洛杉矶分校(UCLA)，计算机系，师从著名的数据库专家Carlo Zaniolo教授，从事数据挖掘领域的合作研究。 2012-2013,挪威科技大学，ERCIM/Marie-Curie Fellow。

目　　录

目录
深度学习基础篇
第1 章绪论 ·································································································· 2
1.1 引言 ······································································································· 2

目 录 深度学习基础篇 第1 章 绪论 ·································································································· 2 1.1 引言 ······································································································· 2 1.1.1 Google 的深度学习成果 ···························································· 2 1.1.2 Microsoft 的深度学习成果························································· 3 1.1.3 国内公司的深度学习成果 ························································· 3 1.2 深度学习技术的发展历程 ···································································· 4 1.3 深度学习的应用领域 ············································································ 6 1.3.1 图像识别领域 ············································································· 6 1.3.2 语音识别领域 ············································································· 6 1.3.3 自然语言理解领域 ····································································· 7 1.4 如何开展深度学习的研究和应用开发 ················································· 7 本章参考文献 ······························································································ 11 第2 章 国内外深度学习技术研发现状及其产业化趋势 ······························· 13 2.1 Google 在深度学习领域的研发现状 ·················································· 13 2.1.1 深度学习在Google 的应用 ······················································ 13 2.1.2 Google 的TensorFlow 深度学习平台 ······································ 14 2.1.3 Google 的深度学习芯片TPU ·················································· 15 2.2 Facebook 在深度学习领域的研发现状 ·············································· 15 2.2.1 Torchnet ···················································································· 15 2.2.2 DeepText ··················································································· 16 2.3 百度在深度学习领域的研发现状 ······················································· 17 2.3.1 光学字符识别 ··········································································· 17 2.3.2 商品图像搜索 ··········································································· 17 2.3.3 在线广告 ·················································································· 18 2.3.4 以图搜图 ·················································································· 18 2.3.5 语音识别 ·················································································· 18 2.3.6 百度开源深度学习平台MXNet 及其改进的深度语音识别系统Warp-CTC ····· 19 2.4 阿里巴巴在深度学习领域的研发现状 ··············································· 19 2.4.1 拍立淘 ······················································································ 19 2.4.2 阿里小蜜——智能客服Messenger ········································· 20 2.5 京东在深度学习领域的研发现状 ······················································· 20 2.6 腾讯在深度学习领域的研发现状 ······················································· 21 2.7 科创型公司（基于深度学习的人脸识别系统） ······························· 22 2.8 深度学习的硬件支撑——NVIDIA GPU ············································ 23 本章参考文献 ······························································································ 24 深度学习理论篇 第3 章 神经网络 ························································································· 30 3.1 神经元的概念 ······················································································ 30 3.2 神经网络 ····························································································· 31 3.2.1 后向传播算法 ··········································································· 32 3.2.2 后向传播算法推导 ··································································· 33 3.3 神经网络算法示例 ·············································································· 36 本章参考文献 ······························································································ 38 第4 章 卷积神经网络 ················································································· 39 4.1 卷积神经网络特性 ················································································ 39 4.1.1 局部连接 ·················································································· 40 4.1.2 权值共享 ·················································································· 41 4.1.3 空间相关下采样 ······································································· 42 4.2 卷积神经网络操作 ·················4

显示全部信息

前　　言

序言1
如今，深度学习是国际上非常活跃、非常多产的研究领域，它被广泛应用于计算机视觉、图像分析、语音识别和自然语言处理等诸多领域。在多个领域上，深度神经网络已大幅超越了已有算法的性能。
本书是深度学习领域的一本力作。它对深度神经网络尤其是卷积神经网络进行介绍，且注重深度学习的实际应用。而且，本书还对深度学习研发现状进行总结和阐述，包括Google 和Facebook。本书通过示例的方式详解深度学习的具体应用，包括手写数字识别，物体识别，及以人为中心的计算（包括人脸识别、人脸表情识别、年龄估计、人脸关键点定位等）。本书也介绍了深度学习Caffe 和Pylearn2 两个平台，并给出具体示例，介绍如何使用。
本书的所有实验均可重现，对初学者、研究生和工程师有重要参考价值，能够帮助读者掌握深度学习的实战技能。我在访问河南大学期间与本书作者相识。作者和他的学生为本书的出版投入了1 年多的时间，付出了巨大的心血。我相信，本书将会被中国科技界所认可。
Ioannis Pitas

序言1 如今，深度学习是国际上非常活跃、非常多产的研究领域，它被广泛应用于计算机视觉、图像分析、语音识别和自然语言处理等诸多领域。在多个领域上，深度神经网络已大幅超越了已有算法的性能。 本书是深度学习领域的一本力作。它对深度神经网络尤其是卷积神经网络进行介绍，且注重深度学习的实际应用。而且，本书还对深度学习研发现状进行总结和阐述，包括Google 和Facebook。本书通过示例的方式详解深度学习的具体应用，包括手写数字识别，物体识别，及以人为中心的计算（包括人脸识别、人脸表情识别、年龄估计、人脸关键点定位等）。本书也介绍了深度学习Caffe 和Pylearn2 两个平台，并给出具体示例，介绍如何使用。 本书的所有实验均可重现，对初学者、研究生和工程师有重要参考价值，能够帮助读者掌握深度学习的实战技能。我在访问河南大学期间与本书作者相识。作者和他的学生为本书的出版投入了1 年多的时间，付出了巨大的心血。我相信，本书将会被中国科技界所认可。 Ioannis Pitas IEEE Fellow 亚里士多德
序言2 深度学习是当前最炙手可热的研究领域，可是缺乏一本系统深入的著作。今天，这本书来了！ 本书是目前国内为数不多的深度学习原著之一，更是最接地气的、同时兼顾深度学习理论与应用开发实践的深度学习著作。本书面向深度学习领域的研究人员、工程师、爱好者，能够帮助人们把握国内外深度学习的研究现状和产业化趋势，快速理解深度学习的基础理论，牢固掌握深度学习应用开发的实用技术。无论是本书的深度学习理论介绍还是应用开发实例详解，作者都采用实例驱动的方式，使得抽象的理论变得容易理解、难懂的图像处理与模式识别问题变得具体、落地。 尤其难能可贵的是，本书还对深度学习平台Caffe 的源代码进行了精彩解析。 本书的10 余个实验都经过作者的反复验证，只需按照书中的相关步骤进行实验，读者就可以重现相关的实验结果。通过本书的学习，读者不但能掌握深度学习相关的基础理论，更能切实掌握深度学习的应用开发技能。因此，对于有意从事深度学习研究与应用实践的研究人员、工程师、爱好者，本书是首选之作。 教授 中国信息协会大数据分会副会长 中国最畅销的《云计算》书籍作者

显示全部信息

媒体评论

在线试读部分章节

深度学习：原理与应用实践下载

深度学习：原理与应用实践

发布书评

相关图书推荐